Background: Etoposide-platinum doublet therapy has been the American standard first-line management for patients with small cell lung cancer for more than a decade. In the 1990s, Japanese investigators developed a regimen of irinotecan-cisplatin for extensive-stage patients and, in one head-to-head trial, the irinotecan therapy produced a statistically significant survival advantage compared with etoposide and cisplatin. We were interested in how American oncologists integrate clinical trial data into their current prescribing practices and studied case-based prescribing recommendations as a surrogate for current clinical practice. We evaluated Network for Medical Communications and Research data from 7 years of live physician meetings to assess changes in use patterns for irinotecan cisplatin since the initial report of the Japanese clinical trial.
Methods: Data from 38 meetings involving 2174 medical professional attendees (MD; DO; MD PhD) were reviewed. Attendees were given a case scenario and asked to select the treatment option that they felt was most appropriate for the hypothetical setting provided. At each meeting, responses were collected electronically and immediately displayed to the attendees. Aggregate response data are held in the Network for Medical Communications and Research database.
Results: During the 7-year study period, regimens including etoposide and platinum have consistently been the most frequently recommended for small cell lung cancer therapy by the meeting participants. The selection of irinotecan-cisplatin by American oncologists was initially infrequent, with a modest, transient impact of a plenary session presentation of the Japanese phase III trial data at the 2000 meeting of the American Society of Clinical Oncology. Publication of the phase III trial findings in the New England Journal of Medicine in early 2002 stimulated a marked increase in irinotecan-cisplatin prescribing with a commensurate decrease in the selection of etoposide and cisplatin therapy. Since 2002, the prescribing patterns have shown an accelerating return to the use of etoposide and platinum despite continuing study of irinotecan-platinum regimens.
Conclusions: Statistically significant advantages demonstrated in a single phase III clinical trial may have only a modest and transient impact on American oncologists' prescribing behaviors. Factors other than the phase III trial results themselves play a role in the likelihood that prescribing behavior will evolve over time. Corroborating trial data among unambiguously relevant populations will be necessary to stimulate a change in standard treatment paradigms.